JPH071904B2 - Opponent response detection method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of use] The present invention is a method for reliably detecting a partner's response in a billing type terminal device connected to a telephone line to which a signal indicating the response of the other party does not arrive. It is about.

[Conventional technology]

Generally, dialing is done via public telephone lines,
When the other party responds, a signal indicating the other party's response such as a reversing pulse is sent from the exchange, and this signal comes from the telephone line. It is said that billing is performed according to the signal.

On the other hand, in some foreign countries, even if the telephone line to which the public telephone is connected, there is one in which a signal such as a reversal pulse does not arrive at the time of the other party's response. In this case, since charging is impossible, the response of the other party is detected by a special means.

That is, the transmitter on the caller side is inactive,
A means for detecting the other party's voice by an amplification detector, etc., and a means for judging the other party's response, or a means for similarly detecting the ringing tone and judging it as the other party's response according to the lapse of a certain time from the end of detection, Alternatively, a combination of the two is commonly used.

[Problems to be solved by the invention]

However, the signals coming from the telephone line include the other party's voice, ringing tone, dial tone, busy tone, and intercept (Intercept.) Sounds. If these are detected, it is misjudged as the other party's response. As a result, unreasonable billing will occur, causing a problem of unexpected damage to the user.

[Means for solving problems]

In order to solve the above problems, the present invention is configured by the following means.

That is, in a terminal device connected to a telephone line where a signal indicating the other party's response does not arrive, each unique frequency component contained in the voice coming from the telephone line and various signal tones is detected for each, and the voice and various signal tones are detected. The detection result is obtained from the table in which each data indicating is assigned corresponding to each combination of each frequency component, and when the detection result is noise that does not indicate either voice or various signal sounds, each frequency component of The minimum detection level is raised and detection is performed again.If the detection result indicates voice, it is determined that this state repeats a predetermined number of times or more within a certain period of time, and the response of the other party is detected by this determination. I am trying.

[Work]

Therefore, depending on each combination of detection conditions of each frequency component, the detection result of whether or not it is voice is obtained, and if the detection result with voice is repeated a predetermined number of times or more within a certain period of time, the voice of the other party's response has arrived. The other party's response is detected as a matter.

〔Example〕

Hereinafter, the details of the present invention will be described with reference to the drawings illustrating examples.

Fig. 2 is a block diagram of a public telephone. It is a line switch L ₁ , L ₂ to which a telephone line from an exchange is connected, a fax switch HS ¹ , a receiving circuit (hereinafter, REC) 1, a diode bridge (hereinafter, A speech circuit (hereinafter, TKC) 5 that forms a speech loop and has a transmission amplifier and a reception amplifier via a DB) 2, a power supply circuit (hereinafter, PS) 3 and a forced disconnection circuit (hereinafter, CB) 4 Is connected to this,
A transmitter T and a receiver R of the handset 6 are connected.

In addition, processors such as microprocessors (hereinafter CP
U) and a control unit (hereinafter, CNT) 7 including a memory and the like, a frequency detection circuit (hereinafter, FDT) 8, a coin processing unit (hereinafter, CPS) 9 for sorting and storing coins,
MF that makes a control judgment according to each output of dial key (hereinafter, DK) 10, fax switch HS _2, etc., and generates CPS9, display unit (hereinafter, DP) 12, audio composite frequency (hereinafter, MF) signal Signal generator (MFG) 13, TKC5, and CB4
Etc. are to be controlled.

The MFG13 responds to the output of DK10 and the control of CNT7, and only when the permission signal OK is given from CNT7, DK10
When the MF signal is generated in accordance with the above operation, the control signal CS is given to the TKC5, the MF signal is sent as a dial signal via the TKC5, and the monitoring information is sent to the maintenance center etc. According to the control of 1), dialing is performed to a specific destination such as a maintenance center in the same manner as described above, and then monitoring information is transmitted via the same route by an MF signal.

Meanwhile, PS3, due on the Futsukusuitsuchi HS ¹ by Ofufutsuku, DC loop is closed via the TKC5, follow for communicating constant polarity and Natsuta loop current by DB2, this was charged Yotsute capacitor, the capacitor The terminal voltage of is supplied to each part as the power supply V.

Therefore, in the case of the first loop method, the DC loop is closed when off-hook is performed, and the CNT7 sends the permission signal OK in response to the output of CPS9 due to the insertion of coins, so the dialing by DK10 is performed. It becomes possible and the call can be made freely by the response of the other party.

However, even if the other party answers, the charging signal such as the reversing pulse does not arrive on this telephone line, and in order to detect the other party's response, the telephone line provided via REC1 using a transformer etc. The incoming voice and various signal tones are detected by FDT8 for each frequency component.
Based on the combination of detection conditions for each frequency component
When the detection result of voice or various signal tones is obtained by the table showing the relation between the detection situation combination stored in the memory of T7 and the detection result, when the detection result shows voice, this state is more than the specified number of times within a certain time. CNT7 by repeating
Determines that the other party's voice, then detects the other party's response based on this judgment, controls the CPS9 accordingly, stores coins, and also erases the miute signal MUT for TKC5. When the inoperable state of T is released and the coins in the standby state become insufficient, the CB4 is controlled to open the DC loop for a predetermined time to restore the exchange and forcibly disconnect the call.

According to DP12, while CNT7 sequentially displays the coin input and accumulation status, CNT7 has a self-diagnosis function and performs self-diagnosis according to the output of the fax switch HS ² which responds to off-hook, and the CPS9 When the storage box is full, the coin is jammed, the connection cord is disconnected due to the theft of the handset 6, and the like, the DP 12 responds by displaying the unusable state and transmitting the above-mentioned monitoring information.

The FDT8 has a variable minimum detection level for each frequency component that can be changed by the control signal CL.
By controlling FDT8, the minimum detection level is successively set to a desired value.

In addition, a busy tone generator (hereinafter referred to as BTG) 14 that locally generates a busy tone is provided, and according to the control of CNT7, the BTG14
Sends a local busy tone to the receiver R of the handset 6 via the transformer T to notify the user of the inability to talk.

FIG. 3 is a block diagram of the FDT8, which includes a preamplifier (hereinafter, HA) 21, a fixed attenuator (hereinafter, PAD) 22, a variable attenuator (hereinafter, ATT) 23, and a limiting amplifier (hereinafter, LA) 24. , PAD25,
Selective amplifiers (hereinafter,
SA) 26 _{1 to} 26 ₈ and an amplification detector (hereinafter referred to as AD) 27 _{1 to} 27 ₈ with a built-in Schmitt trigger circuit make up the main system, and these are integrated circuits and PAD2
2, ATT23 the switching circuit such as analog switch (hereinafter, ASW) 28 ₁ ~ 28 ₄ and summer and freely inserted attenuation settings by, ASW28 ₁ ~28 insertion attenuation time ₄ none off 0 dB, Attenuation of 5 dB when only ASW28 ₁ is on, 10 dB when only ASW28 ₂ is on, 20 dB when only ASW28 ₃ is on, and 30 dB when only ASW28 ₄ is on.
Since the dB, minimum detection level (hereinafter, LDL) control signals CL ₁ -CL ₄ from CNT7 ASW28 ₁ ~28 ₄ ON to comply to, by determining the combination corresponds off, each frequency component of FDT8 You can set the LDL for.

When the insertion attenuation of PAD22 and ATT23 is 0 dB, LDL is HA
It is −55 dBV at the input of 21, and LDL is defined by 5 dB step between −55 dBV and −20 dBV by combining each insertion attenuation amount of PAD22 and ATT23, and the following code is added to each LDL under the control of CNT7. Has been done.

On the other hand, various signal tones coming from the telephone line are
Each frequency component in the following table is included as a unique one, and the optimum LDL for each detection has the relationship shown in the following table.

Tasshi, dial tone (hereinafter, DT), ring tone (hereinafter, RBT),
In the case of busy sounds (hereinafter, BT) and intercept sounds (hereinafter, INT) including chime sounds, the appearance status of unique frequency components continues for a predetermined period of time, and RBT and BT are intermittently interrupted at regular intervals. On the other hand, the main frequency component of voice is 350H.
In the range of z to 1566 Hz, the appearance states of these are unspecified, and the detection results of each are obtained according to these conditions.

That is, in this case, while DT is a continuous sound, RB
T is 1s on, 3s off or 2s on, 4s off repeat, BT is
Repeated 0.5 s on, 0.5 s off or 0.25 s on, 0.25 s off, INT always contains a frequency component of 1800 Hz, and this is continuous for 0.2 s or more.

Therefore, in Fig. 3, the voice and the signal coming from the telephone line via REC1 are amplified by HA21, and PAD2
2 and ATT23 provide attenuation according to LDL, then LA2
Limiting the amplitude at 4, provided via the PAD25 each SA26 ₁ ~ 26 _8, extracts each frequency component of by these connexion 350Hz～1912Hz, more than a predetermined level by AD27 ₁ ~ 27 ₈ Each extract Todoroki after amplified by detecting, with each delivery to 8 bits detect Todoroki Df ₁ ~Df ₈ as CNT7 showing a binary, receives the setting of LDL in accordance with the control signal CL ₁ -CL ₄ from CNT7 Cage,
According to the situation combination of the detected forces Df _{1 to} Df ₈ and their durations, the CNT 7 of FIG. 2 uses the table in the memory to make the above judgment and detection.

Table 3 shows the detection force Df stored in the CNT7 memory.
It is the main part of the table showing the relationship between the situation combinations of _{1 to} Df ₈ and the detection results. When the detection status of the detected forces Df _{1 to} Df ₈ is “1” and the non-detection status is “0”, it is detected. Power Df ₁
Is the lowest, and the detected force Df ₈ is the highest, 16 in every 4 bits.
Since a binary code is formed, 256 kinds of codes of 00 to FF can be obtained by combining the situations of detected forces Df _{1 to} Df ₈ , and data 00, 01 to 07 indicating the detection result are used as a table with this code as an address. Stored in memory as.

Therefore, the read address of the table is specified by the detected forces Df _{1 to} Df ₈ and the corresponding data is read, and the voice or each signal tone of RBT, DT, BT, INT, etc. is read according to this data. The detection result of is obtained.

In addition, in Table 3, in addition to based on Table 2, the frequency deviation of the signal sound is also taken into consideration, and the situation combinations of the detected forces Df _{1 to} Df ₈ are determined, and the main occupied frequency band and frequency components of the voice are determined. The same combination is determined according to the appearance situation of, and 00 of the data indicates the detection result as noise.

FIG. 1 is a comprehensive flow chart of the judgment and control status by the CPU in CNT7, which is executed while executing the instructions in the memory and accessing the necessary data to the memory.

In addition, this control and judgment, depending on the user confirms the arrival of DT and operates the DK10, The MF signal is sent from MFG13, and it starts after the dial transmission by this is completed.

That is, first, a low-level signal is also detected, so that “LDL = 7” 101, which is the lowest LDL by sending the control signal CL, is set, and configured in the CPU to count the number of repetitions for each detection result. After performing "each signal tone / voice counter / clear" 102, the CPU is used to specify a fixed time.
After starting this by the "50ms timer set" 103 provided inside, perform "Df capture" 111 for detection power, check "Df input / existence?" 112, and either bit is "1". If it is Y (YES), then “table reference / detection result is obtained” using the table in Table 3 12
1 is performed, the corresponding "sound / voice counter / count up" 122 is performed according to the result, and then "50ms timer / END?" 123 set by step 103 is checked. During), the process returns to step 111, and steps 111 and thereafter are repeated with a period of 5 ms, for example.

Therefore, 10 table lookups were made in 50ms,
If the same state is repeated or continuous during this period, the corresponding counter will count 10 times by step 122, while the corresponding counter will not count up to 10 times and each counter will not count. The appearance frequency of each detection result is obtained from the count value of, and each detection result can be confirmed by determining that the count value is equal to or more than a predetermined value set to 10 or less within 50 ms.

For the reason mentioned above, as step 123 becomes Y,
How much is the count value of each counter corresponding to step 122 to 10?
"INT counter = 10/10?" 132, "RBT counter 6/10?"
133, judge by BT counter 6/10? 134 sequentially,
If step 131 is Y, the process proceeds to "voice detection process" 141, while "INT detection process" 14 is performed depending on Y of step 132.
On the other hand, in step Y of step 133 and 134, in order to detect voice or INT further, step 102 and subsequent steps are repeated.

When all the steps 131 to 134 are N, the noise is being detected. In order to increase the LDL and avoid the noise detection, the LDL is increased by one step by "LDL = LDL-1" 151 and the fax switch HS ^In accordance with the output of ² , "on-hook?" 152 is checked, and step 102 and subsequent steps are repeated through N of this check.

Therefore, the detection result based on the table of Table 3 stored in the memory is voice, and if this state is repeated 6 times or more during the fixed time of 50 ms, the other party's response is judged by this judgment. Detection is performed, and corresponding processing is executed in step 141.

FIG. 4 shows a lower routine of the “voice detection processing” 141, which performs “release of transmitter operation suspension” 201 by stopping transmission of the miute signal MUT to allow mutual call, and then CPS9.
Control to carry out "coin storage" 202 and allow data service requests by sending MF signals.
Message is transmitted and "data service request permission" 211 is carried out, and then "second DT processing" 212 is carried out in order to prevent chained repetitive dial transmission called chain dial, and step
Similarly to 152, "on-hook?" 213 is checked, and a series of control is ended according to Y of this.

FIG. 5 is a lower routine of the “second DT process” 212. In some foreign countries, the DT is restarted by the on-hook of the other party.
Is sent as the second DT. In this case, it is possible to make a dial call using a hand-held MF signal dial oscillator without repeating on-hook and off-hook. These processes are performed for the purpose of

That is, based on Table 2, “LDL = DT detection level” 301 is set by sending the control signal CL, and the detection outputs Df ₁ , Df ₂ , Df are set.
350, 420, 620Hz is extracted by ₄ and "DT input / Yes?"
Depending on Y of 302, the time when the exchange does not respond even if an MF dial signal is given is judged by "DT input ≥ 20 ms?" 303, and the "100 ms timer set" 311 of the time not affecting the call is set. After that, the transmitter stops the operation of the transmitter 312 by sending the miute signal MUT, and shuts off the input signal from the transmitter T to prevent erroneous detection. ? ”313 is judged, and Y of this is presupposed,“ 100 ms ”set by step 311
While TIMER END? ”314 is N, steps 313 and subsequent steps are repeated.

Further, when the step 313 becomes N, "transmitter operation stop release" 321 is performed by stopping the transmission of the miute signal MUT, and "100 ms tire clear" 322 is performed corresponding to the step 311.

On the other hand, even if the input signal from the transmitter T is cut off, the display data is sent to the DP12 and the message "please handset" is displayed as the DT detection state continues for 100 ms and step 314 becomes Y. 331 to control BTG14 and
After was rows summer against the delivery "332 receiver R, output by the" Onfutsuku? "333 Futsukusuitsuchi HS ² finishes a series of control according to the Y.

Therefore, the second DT arrives according to the on-hook of the other party.
Is detected in response to a continuation of 20 ms or more, and the transmitter T is kept inactive for 100 ms to reconfirm the second DT, and in response to this, unauthorized redialing is blocked, and The user is notified of the inability to make a subsequent call by sending out the local BT and displaying it on the DP 12.

However, the dead period of the transmitter T is 100 ms,
Even if a call is made, this does not cause any trouble.

FIG. 6 shows a sub routine of “INT detection processing” 142,
In the INT detection process, since it is a talkie response and normal mutual calls cannot be performed, the "20s timer set" 401 configured in the CPU is performed, and the sending of the miute signal MUT to the TKC5 is continued and the "speaker The operation stop continuation "402 is maintained, the display data is sent to the DP12 to perform the" Please handset "display" 411, and then the step 401 sets the "20s timer.
END? ”In response to 412 becoming Y, control BTG14 and handset R
To perform a "local BT sends" 421, "Onfutsuku?" 422 ends the series of control according to the Y by the output of Futsukusuitsuchi HS ^2.

Therefore, according to the judgment of INT in step 132 of FIG. 1, step 411 is displayed for 20 seconds, and then the local BT
Is sent to inform the user that the call is impossible.

However, the frequency at which the frequency component is detected may be determined according to each signal tone and the transmission frequency band of the telephone line, and each timer for fixed time regulation and each set time regulation has a voice and The number of repetitions of each detection state may be set according to the determination condition, as well as set according to the intermittent state of the signal sound and the required time for detection and the time required for control.

In addition, in the configurations of FIGS. 2 and 3, the selection is arbitrary according to the situation, and in FIGS. 1 and 4 to 6, the steps are replaced or equivalent in accordance with the judgment conditions. Various modifications are possible, such as the same even if other steps are replaced or unnecessary ones are omitted.

〔The invention's effect〕

As is apparent from the above description, according to the present invention, the frequency components unique to the voice and various signal tones are detected individually,
Since the detection result is obtained from the table according to the combination of each detection situation, the detection result is accurate, and it is judged that the voice detection result has been repeated a predetermined number of times or more within a certain time. Accuracy is improved, and incorrect billing, etc. can be avoided, and various terminal devices such as public telephones, fax machines, etc. that are connected to a telephone line where a signal indicating the response of the other party does not come and bill with coins, fee cards, etc. A remarkable effect can be obtained. If the above detection result shows neither voice nor various signal tones, the minimum detection level of each frequency component is raised and each frequency component is detected again. Is eliminated, and the other party's response can be accurately detected.

[Brief description of drawings]

FIG. 1 shows an embodiment of the present invention, FIG. 1 is a comprehensive flow chart of judgment and control situation, FIG. 2 is a block diagram of a public telephone, FIG. 3 is a block diagram of FDT, and FIGS.
The figure is a flow chart showing the lower routine of FIG. 1 ... REC (reception circuit), 7 ... CNT (control unit), 8 ...
FDT (frequency detection circuit), 22 ... PAD (fixed attenuator), 23
...... ATT (variable attenuator), 26 _{1 to} 26 ₈ ...... SA (selective amplifier), 27 _{1 to} 27 ₈ ...... AD (amplification detector), 28 _{1 to} 28 ₄ ...... AS
W (switch circuit), L ₁ , L ₂ ... Line terminals, Df, Df _{1 to} Df ₈
…… Detected power, CL, CL _{1 to} CL ₄ …… Control signal.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP 59-53899 (JP, A) JP 60-256259 (JP, A) JP 53-8259 (JP, Y1)

Claims (1)

[Claims] 1. A terminal device connected to a telephone line to which a signal indicating a response from the other party does not arrive, wherein each unique frequency component contained in a voice coming from the telephone line and various signal tones is detected for each, Obtaining the result of the detection from a table in which each data indicating voice and various signal tones is assigned corresponding to each combination of each of the frequency components,
When the result of the detection is noise that does not show either the voice or various signal sounds, the minimum detection level of each of the frequency components is increased to perform detection again, and the result of the detection shows the voice. Is a partner response detection method characterized by determining that this state repeats a predetermined number of times or more within a certain period of time and detecting the partner response based on the determination.